Tests are just booby traps for the other engineers so they don't break your code by mistake.
this post was submitted on 24 Sep 2024
810 points (98.7% liked)
Programmer Humor
19282 readers
700 users here now
Welcome to Programmer Humor!
This is a place where you can post jokes, memes, humor, etc. related to programming!
For sharing awful code theres also Programming Horror.
Rules
- Keep content in english
- No advertisements
- Posts must be related to programming or programmer topics
founded 1 year ago
MODERATORS
Its funny cause its true. I often design tests to be "if a case/enum value is added this test will explode and tell them to add code here"
This is why I like strong type systems with exhaustivity checks
Saaaaame. But sometimes even strongly typed stuff wont break on compilation time
'spose that's true enough
Run it in your head, find the edge cases yourself, fix the bug... weakling.
Or do what I do in real life which is patch in new bugs and even a security flaw or two.
While I know that these days, bugs in code can cause real-world harm (personal info leaks, superannuation records lost, lol google), I find it humorous to think of the equivalent, even worse outcomes in my discipline (chemical/process engineering).
"Didn't do any checks, fuck it, I know this calculation is fire 🔥"
Later: 🔥🔥💥
It’s more: I have routed a few pipes in our test system and it’s now spitting out water known to be contaminated but now should have some extra sprinkles in so it’s fine.
What I’m saying is it’s even worse than didn’t do any checks. It’s willfully ignoring existing checks intentionally.
the energy of a chaotic neutral?
"maybe it'll work, maybe it won't, but it'll be FUN"
or chaotic evil?
"naw. fuck y'all's weekend.
Merging failing tests so everybody else has failing tests and wastes time figuring out why.
Nothing neutral here
I haven’t played DnD in like 20 years. Is “Chaotic Dickhead” an alignment now?
That's basically just chaotic neutral
That's what the pipeline is for. It's not that hard to pinpoint the commit that lead to the errors.
If I rebase my branch with main I do not expect any failing tests. If you waste my time merging shit code, fuck you. Fix your shit.
Unless prod is on fire and the CEO is prowling (even then, I'd argue standards should be maintained)
I don't say this is good practice, you shouldn't even be able to merge to main with failing tests. I've implemented an emergency flag to do this, but I don't want to use it in normal, daily business.
"maybe it'll work, maybe it won't, but it'll be FUN"
Flashback to that Tom Cruise Scientology interview 🤣:
It really is ... Fun.
On a scale of one to translunar orbit, how freaking high was he?
Colombia.
I physically reacted to this post with a combination of disgust, anger, and fear. Do tests. All of the tests. Randomize the order in which your tests run. Cover all branches.
Real programmers test in production.
You mean like this?
They were under a lot of pressure.
They were under a lot of pressure.
It was sink or swim.
No time to cave in
You mean like this?
Or like this?
Fuck yeah
AMERICA 🇺🇲🇺🇲🇺🇲🇺🇲🦅🦅🦅🦅❗❕❕🗣️🗣️🗣️🗣️🗣️🔊🔊🔊🔊🇺🇲🦅🦅🇺🇲🦅🇺🇲🦅🍔🍔🍔🍔🍔🍔 FREEDOM 🗣️🗣️🗣️🗣️🔊🔊🔊🔊🔥🔥🔥🔥🥶🥶🥶🥶🥵🥵🥵🥵🦅🦅🦅🇺🇲🇺🇲🇺🇲🇺🇲🇺🇲🇺🇲🇺🇲🦅🦅🇺🇲❗🇺🇲🇺🇲❗🦅❗🦅❗❕❗❕❕❗❕❕❗🇺🇲🇺🇲🇺🇲🇺🇲🗣️🗣️🗣️🔊🔊🔊🔥🔥🔥🔥 DELIVERING DEMOCRACY TO THE DESERT 🦅🦅🦅🦅❗🔊❗❗❗🔥🇺🇲🇺🇲🔥❗❕🔊✈️✈️✈️✈️✈️💣💣💣💣💣💣💣💣💣☢️☢️☢️☢️☢️💥💥💥💥💥🦅🇺🇲🦅🇺🇲🔥🔥🔥🥶🥶🥶🥵🥵🥵🦅🦅🦅🇺🇲🇺🇲
Users will test, don't waste your energy.
tests are for confirming your code STILL works if someone ever changes something
Oh I trust my code, but I don't trust my coworkers not to break something on the very next commit.
I get a small amount of joy from clicking the "request changes" button and blocking some doofus from merging lazy untested code.
I love going into a PR with 3 approvals already and shitting all over it
Tests? Pfffft. I am the test.
And while I'm here: https://blog.jim-nielsen.com/2024/sanding-ui/
Users are the acceptance testers.
It baffles me when people use flex layout when it's clearly visually a grid layout. Nothing here is flexing with varying element sizes and auto-fill-wrap-break of items.
A colleague of mine prefers flex too. But to me, grid is so much more intuitive and simple.
https://css-tricks.com/quick-whats-the-difference-between-flexbox-and-grid/
Tbh I'm not a web person (more of a backend person) and don't know the recommended practices. display: grid; is a good friend of mine xD
I think using display: grid; as your default is the better default, so you're all set. :)
Why do you need either? Just throw the both in the html
People can pull from my cold, dead hands.
(though I'm usually only using it to display some status just for me and not for external consumption; the UI side can have a JSON if it ever comes to that).
I used to be a full-stack dev, but I've been pure backend for so long now, everything I knew is outdated or deprecated.
everything I knew is outdated or deprecated
Given the way the frontend world seems to work, this means you’ve been backend-only for at least a week lol
Weak code lacks tests
Alt: if strength relies on unity I need to switch to game dev
You can't trust others to not break your wonderful code. Write tests for the regression.
"Tester, c'est douter"
The best way is to try it over and over until it works and then assume it works but then go insane wondering where all the edge case bugs are coming from.
I wrote a test one time.